Construction of Tree-based Forest Management Digital Twin Database with Airborne Laser Surveying

Abstract

Recently, in the forestry sector, the carbon fixation capacity of forests has been highlighted for climate change response. In addition, the need for reliable forest management information is increasing for large-scale forest disaster response such as landslides and forest fires. In order to generate essential data for forest management, horizontal structure surveys, such as tree species and forest type, and vertical structure surveys, such as tree height and diameter at breast height (DBH), must be conducted. In this study, a comprehensive survey method using multisensor airborne LiDAR surveying was introduced for Chiaksan National Park (8.32 km2), which is a natural forest. The forest survey method applied in this study was a two-step method that performed object-based forest type classification using high-resolution orthoimages, and then performed individual tree detection (ITD) for each forest using high-density ALS data. As a result of this study, object-based forest type classification using orthophotos showed a classification accuracy of more than 95% for both coniferous and deciduous trees. In addition, in the ITD of natural forests by forest type, the quality of conifers was good, but the ITD quality was higher than 73%. In this process, a method for generating essential data for tree-based forest management, such as tree height and DBH, was established. In addition, we established a process for calculating the stem volume, biomass, and carbon storage capacity of the extracted trees, and created a total of 18 forest management digital twin databases for all trees in the research area. The tree-based forest management digital twin database for national park natural forests constructed through this study was used for the 2D and 3D visualization of various forest management information as well as for the demonstration construction of a forest management digital twin pilot system. Such a tree-based forest management digital twin can quickly confirm more accurate information necessary for forest management by tree unit, so it is expected to be efficiently utilized for establishing a carbon neutrality transition strategy as well as for simulating forest disasters for the conservation management of forest resources.